Periodic Data Relocation in Multi-Tier Storage via Spectrum Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current data relocation methods in tiered storage systems struggle to accurately predict when data should be relocated based on access frequency patterns, particularly for cyclic and seasonal patterns, leading to potential misjudgments and inefficiencies.
Innovation Solution
A method that determines whether data block access frequency is periodic through spectrum analysis, such as Fourier Transform, and calculates the change cycle to prioritize data relocation, improving the accuracy of data block relocation by considering the cyclic or seasonal nature of access patterns.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If data relocation is performed according to a preset schedule or manually, then the storage system can operate with simple control logic, but the accuracy of predicting when relocation should be performed deteriorates
Solution Approach 1:
The patent applies dynamics by transitioning from static preset schedules to dynamic, adaptive relocation timing. The system continuously monitors access frequency and adjusts relocation decisions based on real-time periodicity detection, allowing the relocation strategy to adapt to changing data access patterns while maintaining automated operation
Solution Approach 2:
The patent implements feedback mechanisms by monitoring access frequency of data blocks and using this information to determine periodicity patterns. This feedback loop enables the system to automatically adjust relocation timing based on observed access behavior, improving prediction accuracy without requiring complex manual intervention
2Speed
If data relocation is performed frequently to ensure data is in optimal storage tiers, then data access efficiency is improved, but storage system productivity deteriorates due to excessive relocation operations
Solution Approach 1:
The patent applies periodic action by detecting periodic patterns in data access frequency and scheduling relocation operations accordingly. Instead of frequent or manual relocation, the system identifies the natural periodicity of data access and performs relocation at optimal intervals, reducing unnecessary operations while maintaining access efficiency
Solution Approach 2:
The patent implements preliminary action by predicting future access patterns based on detected periodicity and proactively relocating data blocks before they are needed. This allows the system to prepare data in advance at optimal times, ensuring quick access when data is needed while avoiding excessive relocation operations
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
Enhances the accuracy of data relocation by correctly identifying periodic access patterns and determining the optimal relocation timing, thereby optimizing storage resource allocation and reducing misjudgments.
Implementation Method 1
determining whether access frequency of a data block in the multi-tier storage system is periodic; in response to determining that the access frequency of the data block is periodic, determining a change cycle of the access frequency of the data block
Data Source
AI summary
Techniques for data relocation involve: determining whether access frequency of a data block in a multi-tier storage system is periodic; in response to determining that the access frequency of the data block is periodic, determining a change cycle of the access frequency of the data block; and determining, based on the change cycle of the access frequency of the data block, priority of relocating the data block in the multi-tier storage system.